mlsanityData to 7 Oct 2026

MoEScore: Mixture-of-Experts-Based Text-Audio Relevance Score Prediction for Text-to-Audio System Evaluation

arXiv 2601.06829 ↗First posted 11 Jan 2026cs.SDcited by 0

On the map: Mixture-of-experts → Multimodal MoE models

Abstract (arXiv)

Recent advances in generative models have enabled modern Text-to-Audio (TTA) systems to synthesize audio with high perceptual quality. However, TTA systems often struggle to maintain semantic consistency with the input text, leading to mismatches in sound events, temporal tructures, or contextual relationships. Evaluating semantic fidelity in TTA remains a significant challenge. Traditional methods primarily rely on subjective human listening tests, which is time-consuming. To solve this, we propose an objective evaluator based on a Mixture of Experts (MoE) architecture with Sequential Cross-Attention (SeqCoAttn). Our model achieves the first rank in the XACLE Challenge, with an SRCC of 0.6402 (an improvement of 30.6% over the challenge baseline) on the test dataset. Code is available at: https://github.com/S-Orion/MOESCORE.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Multimodal MoE models, one of 12 directions in Mixture-of-experts. Builds on 15 of its 19 references that are in the map. First posted on arXiv 11 Jan 2026.

Builds on

15 of 19 references.

References · 15

  1. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  2. 2022Large-scale Contrastive Language-Audio Pretraining with Feature Fusion and Keyword-to-Caption Augmentationcited by 707
  3. 2023AudioLDM: Text-to-Audio Generation with Latent Diffusion Modelscited by 592
  4. 2022BEATs: Audio Pre-Training with Acoustic Tokenizerscited by 378
  5. 2023Make-An-Audio: Text-To-Audio Generation with Prompt-Enhanced Diffusion Modelscited by 323
  6. 2022Diffsound: Discrete Diffusion Model for Text-to-sound Generationcited by 256
  7. 2023Make-An-Audio 2: Temporal-Enhanced Text-to-Audio Generationcited by 90
  8. 2017Deep Voice: Real-time Neural Text-to-Speechcited by 83
  9. 2024AudioLCM: Text-to-Audio Generation with Latent Consistency Modelscited by 21
  10. 2024Advancing Multi-grained Alignment for Contrastive Language-Audio Pre-trainingcited by 14
  11. 2025EnvSDD: Benchmarking Environmental Sound Deepfake Detectioncited by 11
  12. 2025M2D-CLAP: Exploring General-purpose Audio-Language Representations Beyond CLAPcited by 10
3 more
  1. 2025TTA-Bench: A Comprehensive Benchmark for Evaluating Text-to-Audio Modelscited by 7
  2. 2024Challenge on Sound Scene Synthesis: Evaluating Text-to-Audio Generationcited by 6
  3. 2025RELATE: Subjective evaluation dataset for automatic evaluation of relevance between text and audiocited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026MoE Adapter for Large Audio Language Models: Sparsity, Disentanglement, and Gradient-Conflict-Free
  2. 2026UniAE-MoE: A Unified Audio Encoder via Mixture of Experts
  3. 2026AnyAudio-Judge: A Dynamic Rubric-Based Benchmark and Evaluator for Audio Instruction Following
  4. 2026AQAScore: Evaluating Semantic Alignment in Text-to-Audio Generation via Audio Question Answering
  5. 2023Retrieval-Augmented Text-to-Audio Generation
  6. 2026MoST: Mixing Speech and Text with Modality-Aware Mixture of Experts
  7. 2025MECAT: A Multi-Experts Constructed Benchmark for Fine-Grained Audio Understanding Tasks
  8. 2024CLAIR-A: Leveraging Large Language Models to Judge Audio Captions